First successful "Hello World" runs of the CNC

[Carl Fisher]

Put a pen refill into the collet and started to get all of my calibration finished up. These are the first 2 successful hello world runs.

Next step will be to install the home and limit switches to get it to home automatically. After that I'd really like to fab up some sort of touch plate for z-axis touch off since doing it by hand seems a bit pesky.

 

[CrealBilly]

Too cool - I feel a little like a new daddy and the CNC is yours not mine - I had nothing to do with it's birth

 

[Carl Fisher]

By the time I get this thing where I want it, it'll cost about the same as giving birth :icon_scra

 

[Hmerkle]

O.K.
So we will learn from you and not make those mistakes! :rotflm:


Looking at the calendar to set up a "See the CNC" visit!

 

[ehpoole]

Carl,

If you do a Z-axis touch plate, make sure it is a spring loaded design with a non-ferrous metal (copper, brass, or aluminum) for a touch plate as there is nothing worse than crashing a carbide bit into a rigid piece of metal and either chipping or crushing the bit. You may also want to rig up either a piezo or optical detector (or whatever will work with your setup) that can accurately detect the moment a bit makes contact with the touch plate so that calibration still works and the bit can not crash even if some paint or debris prevents an electrical path from forming between the touch plate and bit.

It would probably also be a good idea to pause for just a moment before slowly lowering the bit to ensure all lateral motion and vibration has come to rest if you plan to run any especially small/narrow carbide bits as they are especially prone to breakage from excessive lateral force.

These are just a few of the thought processes that I have turned through over the years as I plan for the day I finally get around to building a proper CNC machine...so far thinking about it is as far as I have gotten! Hopefully done of it might prove useful for your project as well.

 

[Carl Fisher]

Your timing is impeccable. I'm actually doing research on touch off plates now.

I'm still trying to wrap my head around how a spring loaded touch plate can give you .001 resolution. I understand the spring will give if the tool crashes hopefully saving the bit, but if the spring gives in any way during the touch off process, wouldn't that invalidate the touch off process? Just things I've been mulling around in.

A cheap or free one can be made without the spring loading from a piece of copper/brass, some wire and an alligator clip. I know it won't protect the bit from a crash, but it would work as a short term solution until I can order a spring loaded one.

Aside from all that, I need to brush up on my cad/cam use again. It's been a little while since I've loaded up my cad package and I've always been lacking on pathing on the cam side but I'll get there.

 

[ehpoole]

Carl,

The spring would be used to press the plate up against a trapped stop (the plate rests against a hole just a bit smaller than the plate's diameter). The spring will hold it tight against that stop unless enough downward force is applied to displace the plate and spring downward. The actual home position would then be a known offset above the plate (the depth of the stop).

Provided you have a fast detection of the contact event (either poll before each down stepping or drive a hardware interrupt on the microcontroller) then you should detect the contact event without actually displacing the spring unless you advance too quickly... In which case you could have a second routine that detects the loss of contact as the home point.

If you secure a piezo disc (speaker) to a brass touch plate then you should also be able to detect the single sharp pulse at the moment of contact, which could provide you with both some redundancy and the ability to detect successfully even if the bit fails to complete an electrical circuit due to paint or debris buildup.

I imagine that one could even make a non-contact optical solution if an IR beam is focused and passed through a narrow slit while slowly spinning the blade...at a certain point the bit will begin to interrupt that beam and you will suddenly see either a train of pulses or a complete loss of the beam depending upon the bit's geometry, you would then double-check by verifying the beam is restored and uninterrupted within a couple of steps up (exactly how many steps would be within tolerance would depend upon your gantry's stiffness).

 

[mkepke]

Hey Carl - if you get a chance, post a full review will you ? I was ogling the Shapeoko 2 kits recently. They have an attractive entry point, but I wonder if they are really a deal by the time they are ready to work ? Sounds like you had CAD/CAM software already ?

-Mark

 

[Carl Fisher]

It depends on what you want a CNC to do right out of the box. I haven't done any cuts with it yet so I can't speak to it's live performance just yet, but I can give you opinions of the build and overall quality of the machine.

I bought the mechanical kit which didn't include any electronics. It comes with everything you need to build a working X-Y-Z axis framework and you add your own steppers, controller, power supply, wiring, etc... They also have a full version that includes NEMA17 motors and an Arduino based controller. Just add your own spindle. I chose to go with the mechanical only kit because I added some upgrades right from the start including 4-axis NEMA 23 260oz-in stepper kit from Probotix, ACME thread for the Z-Axis, and added a drive shaft in place of the dual Y motors. It's a very configurable machine and easy to modify to fit your needs.

The makerslide setup and v-wheels are a nice design for the linear movement. It makes expansion extremely easy. If you want a bigger CNC, just bolt on more makerslides and you've doubled your size. The tedious part was having to tap your own holes for assembly. It's not hard, just time consuming during the build. But it helps keep their cost down and allows them to sell the slides in various kits beyond the ShapeOko.

Out of the box, the usable footprint is kinda small. About 11" x 11" of cutting space in any one operation but as mentioned above, if you just add more makerslides, you increase your size with little effort. The full footprint is about 20" x 21" in stock form.

The new version comes with double X axis for the gantry which seems to be pretty stiff and open ended frame so you can work long pieces in multiple steps without having to worry if it will fit in the machine. I also like the moving gantry design over a moving table because it stays within it's footprint so you know exactly how much space to dedicate to it which is great if you're in a small shop environment.

The belt system is actually very smooth and very fast. I'm still running it through its paces by running programs without the spindle running so I can get the belts stretched tight and get everything nice and square and ready for real cuts.

I chose a DeWalt 660 for the spindle motor for now. It's loud and I have to rely on an external router speed control to slow it down when necessary, but it was cheap and reliable place to start. Eventually I'll pick up a dedicated spindle that is quieter and can be speed controlled by the CNC controller.

Once I do some actual cuts, I'll post more along with some video.

As for the software, I purchased a CAD program so I could do 3D but if you only want 2D, you can get by with something like Inkscape or any other vector image program and then use either a free CAM software for your tool paths or purchase a dedicated CAM package. Once I get my 4th rotary axis up and running, I'll have to dump about $300 to get a CAM package that can do rotary tool pathing, but for now I'm getting by with 2d and 2.5d testing.

 

[micks]

You might try testing the touch plate in the air, above the table. That way, if the plate doesn't register the tool it won't harm anything.

 

[Carl Fisher]

Did my first real cuts today. I'm working on some custom pen boxes for a friend and I figured a small and simple project like this would be a good starting point.

I have to make 4 different cutting operations per box. Basically 3 pieces at 1/2" thick. A bottom with a pocket, a middle with a full cut through, a pocket and some holes for barrel hinges and then a lid with a pocket. The 4th operation is the engraving of the top of the lid.

Right now my biggest difficulties have been getting my homing switches to behave like I want them to and holding down work pieces that are already pretty close to the finished sizes. I'm going to try some double sided turners tape tomorrow. It's pretty thin and holds well on the lathe so it should hold up to machining at 1/16" deep passes.

My test cut today was just in a piece of particle board I had laying around to make sure my dimensions looked good and the machine was behaving as expected.

 

[Hmerkle]

Did my first real cuts today. I'm working on some custom pen boxes for a friend and I figured a small and simple project like this would be a good starting point.

I have to make 4 different cutting operations per box. Basically 3 pieces at 1/2" thick. A bottom with a pocket, a middle with a full cut through, a pocket and some holes for barrel hinges and then a lid with a pocket. The 4th operation is the engraving of the top of the lid.

Right now my biggest difficulties have been getting my homing switches to behave like I want them to and holding down work pieces that are already pretty close to the finished sizes. I'm going to try some double sided turners tape tomorrow. It's pretty thin and holds well on the lathe so it should hold up to machining at 1/16" deep passes.

My test cut today was just in a piece of particle board I had laying around to make sure my dimensions looked good and the machine was behaving as expected.

Carl, I am not really sure this happened because there is nothing too look at (pictures) :rotflm:

 

[ehpoole]

Carl,

If I might make a suggestion, leave 2-4 tabs about 1/16" by 1/8", or thereabouts, at points around the perimeter (just skip the final 1/16" pass where each tab is to be left). The tabs will hold things in place much more securely than double-sided tape and are much easier to remove without damage - just snip with some pliers, then a one second removal process at a disc or band sander for each tab and there will be no evidence.

For larger pieces you can use an old dust collector motor to draw a vacuum on the table, with holes distributed throughout the table, the vacuum will then help to hold larger pieces stationary if you have movement problems and don't want yo leave tabs on them -- or just use tabs on everything.

 

[Carl Fisher]

Hi Ethan.

Tabs are built into the CAM packages I use. I just grab the tool path I want to add them to and it's just a few button clicks. Keeps everything from binding on the last pass.

The double sided tape is for cutting something that there is no room for clamping. i.e. I have a 7" x 3" pre-cut blank that I need to create a few pockets and engravings but no outside profile cuts. I have a lower plate with some blocks to index the pieces since I have a lot to do but I can't figure any other way to hold them down besides tape. It's faster to cut the blanks to size on the tablesaw than to wait the CNC to do it.

For larger pieces, I'm going to add some t-track to my spoil boards and just use normal hold down clamps clear of the cutting area and tabs like you suggested to hold the cut pieces in place.